Board Meeting Protocol — Business

Runs a structured 6-phase multi-agent board meeting for strategic decisions, preventing groupthink and capturing minority views. It produces a synthesis with a recommended decision and action items, then extracts a founder-approved decision record.

c-level-advisor claude-code

📂 View skill on GitHub ↗ 🚀 How to use

✨ What it does

6-phase protocol: context, independent contributions, critic analysis, synthesis, founder review, decision extraction.
Isolated agent contributions with self-verification and confidence scoring.
Role selection based on topic (e.g., market expansion, product direction).
Adversarial critic analysis to flag suspicious consensus and missing voices.
Two-layer decision memory: raw transcripts and approved decision records.
Human-in-the-loop founder review with override authority.

🎯 When to use it

When the user invokes /cs:boardroom or calls a board meeting.
When a strategic question needs structured multi-perspective executive deliberation.
When you want to avoid groupthink and capture independent C-suite perspectives.
When a decision needs to be documented in a clean, approved record.

🚀 How to use

Trigger: /cs:boardroom

Trigger with /cs:boardroom followed by the strategic question. The skill loads company context and approved decisions, then runs the 6 phases, pausing for founder approval before logging the decision. Example prompts:

/cs:boardroom Should we expand to Spain in Q3?
/cs:boardroom Should we adopt a new AI model for our product?

📄 Output: A synthesis document with recommended decision and action items, plus an approved decision record saved to ~/.claude/decisions/approved/.

📦 Add this skill to Claude Code

# 1. Get the skills repo
git clone --depth 1 https://github.com/alirezarezvani/claude-skills /tmp/claude-skills

# 2. Copy this skill into your project (or ~/.claude/skills for all projects)
mkdir -p .claude/skills
cp -r /tmp/claude-skills/c-level-advisor/skills/board-meeting .claude/skills/board-meeting

Skill source: c-level-advisor/skills/board-meeting/SKILL.md

⚠️ Good to know

Requires company context file at ~/.claude/company-context.md and existing approved decisions directory; founder must be available for Phase 5 review.

❓ FAQ

What happens if the founder disagrees with the recommendation?

The founder can modify or reject the proposal; user corrections override agent proposals with no pushback.

How does the skill prevent groupthink?

Phase 2 runs agents in isolation with no cross-pollination, and Phase 3 critic analysis flags suspicious consensus.

What memory does the skill load in Phase 1?

It loads only Layer 2 approved decisions from ~/.claude/decisions/approved/, never raw transcripts.

🤖 Overview, features, install steps and FAQ were generated from the project's SKILL.md on Sep 4, 2026. Always check the original source before running commands.